Extracting Hardware Reconfiguration Models Based on Knowledge Synthesis from STEP Files

Today's industrial manufacturing challenges force manufacturers to optimize and increase the flexibility of their facilities. In practice, this requires analyzing, preparing for adaptation, and adapting brownfield manufacturing systems. The digital twin, a digital representation of a manufacturing system, is a key enabler for efficiency, flexibility, and sustainability. Unfortunately, the analysis and preparation of brownfield systems for adaptation, as well as the creation of digital twins, are challenging and time-consuming tasks. This paper presents an approach to automatically create digital models of systems based on 3D CAD models. To this end, the CAD data, stored as a STEP file, is analyzed to extract relevant information for a subsequent graph analysis, which is used to identify components, their dependencies, and the resulting functional modules. Finally, the gained knowledge is transformed into feature models that can be used as a digital model for configuration selection to support automatic reconfiguration planning of brownfield manufacturing systems. The developed approach is evaluated based on an industrial use case of a soft gripper system.